Computational Modeling of Brainstem Stimulation

Computational neuron models have become important analytical tools for predicting how membrane properties impact neuronal behavior in well-controlled environments. Here, we develop a novel computational approach to investigate how alterations in the extracellular electric field around such neurons can also influence a cell’s probability of generating an action potential. Specifically, we have focused our efforts on developing computational neuron models within a region of the brainstem known as the pedunculopontine nucleus (PPN), which is thought to be critical in the maintenance of posture and the initiation and modulation of gait. The experimental challenge with stimulation in the vicinity of the PPN is the close proximity to axonal fiber tracts that when stimulated can induce unintended changes in neuronal pathways encoding vestibular, somatosensory, and auditory information. In this study, biophysically-realistic neuron models were developed for two classes of cells in the PPN and three classes of axonal fibers of passage, those within the superior cerebellar peduncle (SCP), medial lemniscus (ML), and lateral lemniscus (LL) pathways. Locations for each cell type were based on histological and magnetic resonance imaging, and were coupled with electrostatic finite element models that simulate the voltage distribution generated during electrical stimulation in the vicinity of the PPN. The results indicate that monopolar electrode configurations will result in poor stimulation selectivity, suggesting a need for more advanced multi-polar stimulation architectures.